Issue setting up Docker Containers with reComputer J4012

Robotics & Edge Computing Jetson & Embedded Systems Jetson Orin NX
1

I want to setup a docker container that runs on my reComputer j4012, based off of the Orin NX 16gb. But I can’t seem to get the container to have access to the GPU, I have tried the standard “–gpus all” and other commands to try and fix this. I have tried building off of the Nvidia L4T ML container, Nvidia L4T Base, and Nvidia L4T Cuda. But I also need pytorch for my use of Meta’s Segment Anything Model.

I have been in dependency troubles for awhile and when I get them to align for my use of SAM, the container doesn’t have access to the gpu.

Can anyone help me or point me in the right direction of a container that should work for my situation?

3

Hi,

On Jetson, please launch the container with --runtime nvidia to enable GPU access inside the container.

For SAM, you can check the below tutorial to enable it on Jetson:

Thanks.

4

Does SAM have to be run through this Jetson container or can I use it how Meta specifies on their repository, downloading it and installing a checkpoint?

5

Hi,

You can install it natively.
But you can refer to the installation steps as the reference:

Thanks.